RE: Problem understanding set-register syntax

["Drew Adams" <drew.adams@oracle.com>]

> > Why can't I just use:
> > (set-register i '(file . "~/.emacs.d/init.el"))
> > ; will say Symbol's value as variable is void:i
> 
> Because the first argument to `set-register' is a register 
> name, which is a character, and `?i' is the Lisp syntax for
> the character `i'.

I should also have mentioned that using just i, without quoting it (i.e., 'i)
causes the symbol i to be evaluated.  And since the symbol i has no value here
(no global value) you get the error message saying that its value is void.

Some symbols, such as t and nil (and all keywords, such as :foobar) are
self-evaluating, meaning that they act just like variables whose values are the
symbols themselves.  You do not need to quote self-evaluating symbols in an
evaluation context.  Other symbols (e.g. i, foobar) you do need to quote, if you
want not the value of the symbol (e.g. it might not have a value) but the symbol
itself.

But all characters are self-evaluating, so ?i evaluates to ?i.

See the Emacs manual, node `Init Syntax'.
See the Elisp manual, node `Basic Char Syntax'.
(Use `g' in Info to go to a particular node.)